Support Vector Regression and Ant Colony Optimization for Pollutant Emission Control in Power Plants
Pollutant emission from the power plant poses terrible threat to the surrounding environment. As air emissions standards become more stringent in China and around the world, coal fired power plants face important challenges concerning the methods and technologies theyll use to meet these new environmental requirements. In the current study, Support Vector Regression (SVR) is proposed to model the functional relation between the Nox emission and operational parameters of a utility boiler, and is also compared with the currently popularly-used neural network-based modeling tool. Subsequently, application of Ant Colony Optimization (ACO) to control pollutant emission is described. A massive of thermal field test data, which downloaded from the actual power plant, is employed to train and validate the SVR model. The technology based on SVR and ACO for pollutant emission is discussed in detail. The results show that SVR, compared to BPNN, can deal with highly nonlinear Nox emission with good accuracy as well as high stability and robustness. The hybrid algorithm by combining SVR and ACO can effectively reduce Nox emission of coal-fired utility boiler. The results can be contributed to pollutant emission reduction to environment in actual power plant.
environment pollutant emission control coal-fired power plant support vector regression ant colony optimization
ZHENG Ligang JIANG Linhua YU Minggao YU Shuijun
School of Safety Science and Engineering, Henan Polytechnic University, Jiaozuo 454000, Henan, China
国际会议
2007环境科学与技术国际会议(The 2007 International Symposium on Environmental Science and Technology)
北京
英文
2007-11-13（万方平台首次上网日期，不代表论文的发表时间）